Research

・Development of 4D-CT based ventilation using deformable image registration for radiotherapy

Lung cancer is the leading cause of cancer, with the highest mortality rate in Japan. At present, surgical resection remains the standard and curative modality for treatment of early stage non-small-cell lung cancer (NSCLC). However, some patients cannot tolerate surgery due to advanced age or comorbidities such as emphysema and heart disease. Thus, radiation therapy plays a key role in the treatment of early stage NSCLC.
Functional avoidance may reduce pulmonary toxicity. Several papers showed a stronger correlation of radiation pneumonitis with ventilation-weighted lung dose than with physical dose. Several studies have been carried out with the treatment planning using lung functional imaging to reduce the dose to the high-functional lung region. The potential of incorporating functional planning with single photon emission computed tomography (SPECT) and magnetic resonance (MR) imaging has been investigated. However, there are several drawbacks; SPECT has low resolutions and aerosol deposition in the airway. MR imaging with hyperpolarized gas is complicated by the need for tracer gas and specialized equipment. New ventilation images created by four-dimensional computed tomography (4D-CT) with deformable image registration (DIR) have been developed (Fig.1). Because 4D-CT data are routinely acquired for lung cancer treatment planning, calculating 4D-CT ventilation image doesn’t add any extra dosimetric and monetary cost to the patient. In our group, we strive to develop this novel ventilation and to prepare this ventilation for clinical practice. Recently, we are scheduling to do the clinical trial using this ventilation imaging (phase II) with collaboration of Hiroshima University (Principal investigator: Tomoki Kumura) (Fig.2). In addition, we are developing the specific QA phantom for deformable image registration (Fig.3).

Grants & Funding:

・Japan Society for the Promotion of Grant-in-Aid for Scientific Research (C) (17K10478)
(April 1, 2017~March 31, 2020、PI: Tomoki Kimura, Co-investigator: Noriyuki Kadoya)

・Research collaboration with Hitachi (July 1, 2018~)

・Clinical Research Fund of Tokyo Metropolitan Government (H290303006) (April 1, 2017~March 31, 2019、PI: Yujiro Nakajima)

・Research grant of the foundation for Promotion of Cancer Research (April 1, 2017~March 31, 2018、PI: Yujiro Nakajima)

・Tohoku University Hospital Grant of Young Investigator in Clinical-Translational Research (April 1, 2015~March 31, 2017、PI: Noriyuki Kadoya)

・Research collaboration with Chiyoda Technol Corporation (April 1, 2014~)

・Research grant of the Japan Radiological Society from Bayer (April 1, 2014~March 31, 2015、PI: Noriyuki Kadoya)

・Japan Society for the Promotion of Science-in-Aid for Young Scientists (B) (24791268) (April 1, 2012~March 31, 2014、PI: Noriyuki Kadoya)

・MRI-Linac

Magnetic resonance imaging (MRI) is a medical imaging technique used in radiology to form pictures of the anatomy and the physiological processes of the body in both health and disease. MRI scans are painless and provide greater detailed images of soft tissue than CT scans. With the advance of MRI technology, several research teams (companies) can develop the MR-Linac system. Merging an MRI with a linear accelerator allows greater precision in cancer treatment. In our group, we strive to develop novel dose calculation algorithm and treatment planning method for MR-Linac. In addition, we are developing and evaluating deformable image registration (MR-MR/MR-CT).

Grants & Funding:

・Japan Society for the Promotion of Science-in-Aid for Young Scientists (B) (18K15617) (April 1, 2018~March 31, 2020、PI: Kengo Ito)

・Japan Society for the Promotion of Science-in-Aid for Young Scientists (B) (15K19198) (April 1, 2015~March 31, 2017、PI: Kengo Ito)

・Research collaboration with MIM (April 1, 2017~)

・Automation of radiotherapy process using A.I. Artificial Intelligence

The advancement of treatment modalities such as IMRT and VMAT has resulted in good outcomes because of dose concentration. On the other hand, it has also resulted in increased plan complexity, decreased consistency of plan quality and increasing increased time. To tackle this issue, we focus on A.I. to automate the various parts of radiotherapy process (e.g., segmentation, treatment planning, and patient specific QA). Recently, we could develop the automation system for predicting the dose distribution and patient specific QA.

Grants & Funding:

・Research grant of the foundation for Promotion of Cancer Research (April 1, 2018~March 31, 2019、PI: Yoshiki Takayama)

・Sendai Medical Center Research grant (April 1, 2018~March 31, 2019、PI: Seiji Tomori)

・Development of prediction system for clinical outcome and treatment response using huge patient’ data and image features (Radiomics study)

We strive to develop novel prediction system for clinical outcome and treatment response using bigdata and Radiomics technique. Radiomics is a field of medical study that aims to extract large amount of quantitative features from medical images. These features have the potential to uncover disease characteristics that fail to be appreciated by the naked eye. This technique is useful for predicting prognosis and therapeutic response for various conditions, thus providing valuable information for personalized therapy. In our group, we are developing the radiomics-based prognosis prediction system in lung and esophagus cancer patients. In addition, we are developing the novel radiomic feature to improve the radiomic analysis.

Grants & Funding:

・Research collaboration with Osaka University Graduate School of Medicine (April 1, 2018~)

・Development of accurate and precise methods for dose accumulation with novel deformable image registration algorithm for cervical cancer patient

In general, external beam radiotherapy (EBRT) with Linac and high-dose-rate brachytherapy (BT) with RALS were used for treatment of cervical cancer (Fig.2). As the two treatment plans use different planning CT images, it is impossible to add two dose distributions. To solve this issue, deformable image registration (DIR) technique enable us to make dose accumulation between dose distributions based on different CT images. However, due to presence of intracavitray applicator in clinical use of BT, the DIR between BT and EBRT is a challenging (Fig.3). To tackle this problem, our group focus on development of novel DIR algorithm that achieves accurate DIR accuracy even though deformation is considerably large.

Grants & Funding:

・Japan Society for the Promotion of Science-in-Aid for Young Scientists (B) (15K19199) (April 1, 2015~March 31, 2018、PI: Noriyuki Kadoya)

・Research grant of the Japanese Society for microSelectron HDR (April 1, 2014~December 31, 2015、PI: Noriyuki Kadoya)

・Research grant of the foundation for Promotion of Cancer Research (April 1, 2014~March 31, 2015、PI: Yusuke Onozato)

・Establishment of standard protocol for DVH-based patient specific QA for IMRT

Gamma index evaluation has become a standard technique used to compare measured distributions with calculated distributions by a commercial radiation treatment planning system (TPS). A typical example of an acceptance criterion of 95% of points above a dose threshold must have a gamma index less than one for dose difference and distance–to-agreement limits of 3% and 3 mm, respectively. Previous studies demonstrated a lack of correlation between conventional IMRT QA methods and dose errors in anatomic regions-of-interest and reported that the gamma passing rate has a weak correlation to critical patient dose volume histogram (DVH) errors.
To solve this issue, there has been a growing interest in 3D validation of treatment delivery with IMRT. For 3D validation, dose reconstruction can be performed by several techniques, such as measurement-guided technique with commercially device ArcCHECK (Sun Nuclear Corporation, Melbourne, FL, USA) or log file-guided technique. The log file-guided technique provides several advantages that a series of procedure in patient-specific QA can be performed with “device-free”, and requires only electric file, so-called log file. We strive to determine the accuracy of two techniques and to establish a standard protocol for DVH-based patient specific QA.

Grants & Funding:

・Research collaboration with Sun Nuclear

・Research grant of the foundation for Promotion of Cancer Research (April 1, 2018~March 31, 2019、PI: Yoshiyuki Katsuta)

・Research grant of the foundation for Promotion of Cancer Research (April 1, 2016~March 31, 2017、PI: Yoshiyuki Katsuta)

・Clinical applications of 3-dimensional printing in radiotherapy

3D printing had become very popular over the years, and several papers about clinical applications of 3D printing in radiotherapy have been already published. In our group, we strive to develop the novel anthropomorphic head and neck phantom with tissue heterogeneity using 3D printer (Fig.14-15). In addition, we created the deformable cervical phantom to do the DIR QA. We believe that 3D printed phantom can be suitable validation tools in various parts of radiotherapy process.

Grants & Funding:

・Research collaboration with Innovation gate

・Research collaboration with Tetraface

・Development of CBCT-based adaptive radiotherapy

Recent studies have shown that not only tumor size and organ position change throughout the course of radiation therapy, but organ size and density also change. This warrants the need for adaptive radiation therapy based on cone beam computed tomography (CBCT) installed on the linear accelerator. However, establishing a standard protocol for CBCT-based dose calculation is not straightforward due to difference in HU values between planning CT and CBCT (Fig.4). Especially, for proton therapy, the lower quality of CBCT image and high scatter-to-primary photon ratio may render the CBCT image unsuitable for direct proton dose calculation. To tackle this problem, we strive to develop a HU modification method and vertical planning CT to improve the CBCT-based dose calculation accuracy.

・4D dose calculation for dynamic MLC-based tumor tracking on radiotherapy for lung cancer

Our long-term goal is to develop an adaptive treatment planning and delivery system to account not only for tumor motion but also for its deformation using a markerless dynamic MLC-based tumor tracking method in radiotherapy for lung cancer. We strive to determine the dosimetric impact of the tumor tracking radiotherapy plan (tracking plan) compared with the respiratory gated radiotherapy plan (gated plan) and continuous 4D-CT plan (non-gated plan) using 4D treatment planning (Fig.5). Furthermore, we aim to establish a standard treatment planning strategy for dynamic MLC-based tumor tracking.

Division of Medical Physics, Department of Radiation Oncology, Tohoku University Graduate School of Medicine,
1-1 Seiryo-cho, Aoba-ku, Sendai, 980-8574, Japan MAP
Tel:+81-22-717-7433 Email: kadoya.n(at)rad.med.tohoku.ac.jp